摘要
随着无人机在军民领域的大规模应用,对无人机复杂工况下的安全性和适应性提出更高要求。闭环辨识能够在考虑系统反馈作用下获取辨识参数,大幅提高辨识试验的稳定性和安全性。针对无人机闭环辨识试验中多变量非线性气动参数获取问题,首先基于某固定翼无人机动力学模型,推导了无人机的气动力、气动力矩等29个气动参数辨识模型;然后提出了一种闭环激励下无人机多变量非线性参数频域在线辨识方法,设计了基于递推傅里叶最小二乘的闭环辨识的流程架构;最后通过添加噪声后的仿真系统对辨识方法的有效性进行了验证,对辨识结果的有效性进行了分析,可为后续开展的闭环辨识飞行试验提供技术储备和理论支持。初步仿真结果证明,提出的闭环激励下无人机多变量非线性参数频域在线辨识方法可以高效获得辨识结果,辨识的29个气动参数中有16个气动参数辨识平均误差3.43%,同时提出的辨识方法较常规最小二乘法精准度平均提高25.5%.
With the large-scale application of UAVs in the military and civilian fields,higher requirements are placed on the safety and adaptability of UAVs under complex working conditions.Closed-loop identification can obtain identification parameters when considering system feedback,which can greatly improve the stability and safety of identification tests.Aiming at the problem of obtaining multivariable nonlinear aerodynamic parameters in the UAV closed-loop identificatioii test,firstly,based on the dynamic model of a fixed-wing UAV,29 aerodynamic parameter identification models such as the aerodynamic force and aerodynamic moment of the UAV are derived;then,a frequency-domain online identification method for multivariable nonlinear parameters of UAV under closed-loop excitation is proposed,and a closed-loop identification process framework based on recursive Fourier least squares is designed;finally,the identification method is verified by the simulation system with adding noise,the validity of the identification results is also analyzed,which can provide technical reserves and theoretical support for the subsequent closed-loop identification flight test.Preliminary simulation results prove that the proposed online identification method for multivariable nonlinear parameters of UAVs in the frequency domain under closed-loop excitation can efficiently obtain identification results,and the average error of identification for 16 of the 29 aerodynamic parameters identified is 3.43%;The accuracy of the proposed identification method is increased by 25.5%on average compared with the conventional least squares method.
作者
蒋铁英
JIANG Tieying(Aerospace Times Feihong Technology Co.,Ltd.,China Academy of Aerospace Electronics Technology,Beijing 100094,China;Intelligent Unmanned System Overall Technology Research and Development Center,China Aerospace Science and Technology Group Co.,Ltd.,Beijing 100094,China)
出处
《无人系统技术》
2023年第2期32-41,共10页
Unmanned Systems Technology
关键词
闭环激励
无人机
非线性参数
在线辨识
傅里叶变换
最小二乘
频域变换
气动参数
Closed Loop Excitation
Unmanned Aerial Vehicles
Nonlinear Parameters
Online Identification
Fourier Transform
Least Square
Frequency Domain Transformation
Aerodynamic Coefficient
